Longitudinal Study of Children's Sensory Response Patterns: Stability, Change and Treatment Effects

Background: Sensory symptoms, which were recently added to the DSM-5 diagnostic criteria, are highly prevalent in autism spectrum disorder (ASD) (Baranek et al., 2014; Lane et al., 2014; Ben-Sasson, 2008). Sensory response patterns (e.g., hyperresponsiveness, hyporesponsiveness, sensory seeking) are known to impact functional outcomes (Watson et al., 2011; Boyd et al., 2010) and are often targets of intervention. Although cross-sectional research with children suggests differences across age, prospective longitudinal studies of sensory response patterns are lacking, particularly those combining parent report and observed measures.

Objectives: (1) To determine the stability of sensory response patterns from early to later childhood in children with ASD in comparison to those with other developmental disorders (DD). (2) To determine if dosage of traditional therapies (i.e., Occupational Therapy [OT], Physical Therapy [PT], Speech-Language Therapy [SLP]) impacts sensory response pattern scores via main effects or through interactions with earlier sensory scores for children with ASD.   

Methods: Children with ASD (n=57) and DD (n=34) ages 2-12 years participated in this longitudinal study.  Four measures –two parent report (Sensory Experiences Questionnaire & Sensory Profile) and two observed (Sensory Processing Assessment & Tactile Defensiveness and Discrimination Test-Revised) –were collected at time 1.  At time 2 (~3-5 years later) parent-reported sensory measures were collected and structured interviews detailing treatment history between time points were conducted. Treatment was coded for type, and average monthly dosage for OT, PT, and SLP was calculated.  General linear models were run separately for the two parent report measures (combined) and the two observed measures (combined) at time 1 to predict time 2 outcomes, for each of the three sensory response patterns. An interaction term for diagnosis (ASD/DD) and covariates (child’s age, IQ, gender, time between assessments, mother’s education and income) were included. Treatment dosage was subsequently added to determine treatment effects, including interactions with time 1 scores.

Results: Mean sensory symptoms decreased over time, significantly for hyporesponsiveness in the ASD and DD groups (both p<.05), and for sensory seeking in the ASD group (p=.01). Changes in hyperresponsiveness did not reach significance. The time 1 parent-reported scores were strong predictors of time 2 parent-reported scores for all three constructs (partial r= .65-.73, all p<.001). The observed scores did not show strong associations to the later parent-reported scores. There was evidence for treatment main effects on all three sensory response pattern scores at time 2 (all p<.05), but in the opposite direction predicted (i.e., higher dosage of traditional therapies between time 1 and 2 was associated with higher sensory scores at time 2). There was no evidence for treatment moderation of time 1 scores on time 2 outcomes.

Conclusions: All three sensory response patterns were strongly stable from early to later childhood when using parent-reported measures. Because treatment was not manipulated in this study, causal inference is limited; however, we hypothesized that children with more severe sensory problems required and/or received more treatment, rather than inferring that treatment led to higher sensory symptoms over time. OT and SLP (not PT) appeared to contribute most to these findings.